add tokenizer.save_pretrained(MODEL) right under the model's save_pretrained
tokenizer.save_pretrained(MODEL)
Β· Sign up or log in to comment